|
Post by johneppstein on Apr 2, 2020 20:40:00 GMT -6
If you have a preamp, like a 1073, with input trim and an output attenuator, its good practice to leave the output atten at unity (all the way up), set the input trim till you reach your desired level into the DAW, and use the output atten to dial or fine tune the level if you still need to. BUT, if you want to push the preamp into blishful distortion, you keep pushing the input trim up till you hear the preamp breakup in a pleasing way, and keep pulling the output trim down to compensate. I usually keep my peaks between -12 and -6. At 24bit the only danger in recording too hot is clipping your converters, usually with fast transients (like drums) or really dynamic sources (like some vocalists). Even then, just because the red light hits intermittently doesn't necessarily mean you've yet clipped your converters. If I'm not mistaken, the original 1073 did not have an output attenuator. The original 1073 was a channel in a console. Those have these things called "faders". In most cases the "output attenuator" on a standalone pre is essentially the same thing. A pot's a pot.
|
|
|
Post by johneppstein on Apr 2, 2020 20:43:56 GMT -6
No audio gear that I know of has specs for compression(saturation or clipping) point. They never tell us at what level their gear will start to clip. IF they did that, then you might start to be able to devise a mathematical way to determine optimal levels, but until then, it's just finding it for yourself through the "listen for the crackle" method. Yes, because it's a moving target depending on the type of circuit, the kind of distortion it makes, and what the knee of that distortion is, depending on the amount of negative feedback applied. You can spec 1% distortion but that tells you nothing about the harmonic content, or where 3% is, or 5%, or 10%. Some things it's another 0.5dB, others it's another 10dB. "specs", schmecks. MEASUREMENTS!
I never take published specs as much more than a rough guideline. Even a fairly detailed spec sheet leaves too much out (and the average customer probably couldn't make much sense of it anyway.)
|
|
|
Post by chessparov on Apr 2, 2020 20:51:49 GMT -6
God invented the oscilloscope. Then had them Made In China.
This was all before the original Cinemascope BTW. Chris
|
|
|
Post by johneppstein on Apr 2, 2020 20:57:21 GMT -6
Yes, because it's a moving target depending on the type of circuit, the kind of distortion it makes, and what the knee of that distortion is, depending on the amount of negative feedback applied. You can spec 1% distortion but that tells you nothing about the harmonic content, or where 3% is, or 5%, or 10%. Some things it's another 0.5dB, others it's another 10dB. I know this. That's why I'm saying it. In RF, there's a point called P1dB. It's the point of incident signal compression equal to 1dB. In layman's terms, you run out of voltage headroom in the amplifier, so the signal no longer gets louder. If you were to increase the input signal beyond the saturation point so that the output can no longer increase by 1:1 ratio, there will be an amount of signal on the input that equates to the output not being able to increase by 1dB. P1dB doesn't care about the types of harmonics created by the saturation, because as you mentioned, it's highly dependent on the types of circuits, but it's a measurement that could easily be used for all amplifiers because it defines the point of nonlinearity that can be used to determine the linear range for any amplifier. I think the problem is that most audio devices fudge their headroom numbers a little, and using compression point would probably dispel some of the marketing fluff they rely on. We call the point where we run out of voltage headroom "clipping". Strictly speaking it's not really the same thing as "saturation", although it's a related phenomenon. If one wishes to get nitpicky, saturation refers to the point in transformers/inductors where the core can no longer handle an increase in magnetic field; it becomes magnetically saturated, like a sponge that can no longer absorb additional water. It's not when the "water" supply runs out.
|
|
|
Post by johneppstein on Apr 2, 2020 20:58:45 GMT -6
"Set the Attack" then..."Release the Crackle!" (love them old stop motion Animation Movies! Chris Don't forget the "Snap" and the "Pop"!
|
|
|
Post by johneppstein on Apr 2, 2020 21:09:38 GMT -6
Intersample peaks are a very big deal John but no converter has a noisefloor of -144 dB. The analog parts of it don’t. The switchers in the chip don’t. The mics and mic pres don’t. You’re hitting resistor and switcher electrical noise. And all of that stack up of external and gear noise is just being converted and saved to the file in the computer. The converter outputs it as a 24-bit file but it itself has a lesser dynamic range. A cheap, crappy converter like a Focusrite Scarlet, Presonus, or Steinberg not the AXR4 thing, can’t even hit 16-bit THD+N. A workman like Babyface Pro will measure slightly cleaner than redbook standard. Now that’s just with a 1 kHz sine wave. When used with material that isn’t fixed sine waves, all bets are off. A Babyface Pro sure as hell isn’t that clean in the high end. The RME treble (not as bad as before but still bad) won’t allow that. Stack a bunch of tracks through it and the noise and distortion compound. Hence the constant search for cleaner and more accurate conversion or the acquiescence to conversion that distorts in a pleasing way like the Lavry Saturation and Burl transformers. We're talking anbout dBfs. That's a digital measurement; it has no bearing in the analog realm. A well designed converter should have an analog section that can deal with that. And what you furnish to the ME these days is a digital file; it shouldn't be affected by the analog output stage of your converter. The file that leaves the ME won't (or shouldn't) be at the same level as the input - part of mastering includes optimizing the output level for the delivery platform.
|
|
|
Post by johneppstein on Apr 2, 2020 21:16:27 GMT -6
It’s the Zen of audio balancing noise, overload, and tone. You do what you have to reach your own sonic nirvana.
Some noise can be nice.... With the years passing by....I have my problems with all those gain staging theories....
Pushing the input like crazy in context A can sound wonderful, abusing the same gear in context B sounds awful. A wise man on this MB told me "Use your ears" ( svart )
Note that pushing the input to your preamp/analog input chain does NOT mean that you need to push the input to the converter. There are these things called "output level controls" (aka "faders" in some contexts.) They're there for a reason.
|
|
|
Post by ml on Apr 3, 2020 6:58:20 GMT -6
What is the proper way to gain stage an LA-2A? I leave it in the chain with the reduction all the way down and the gain at 40 to start.
|
|
|
Post by svart on Apr 3, 2020 8:09:19 GMT -6
I know this. That's why I'm saying it. In RF, there's a point called P1dB. It's the point of incident signal compression equal to 1dB. In layman's terms, you run out of voltage headroom in the amplifier, so the signal no longer gets louder. If you were to increase the input signal beyond the saturation point so that the output can no longer increase by 1:1 ratio, there will be an amount of signal on the input that equates to the output not being able to increase by 1dB. P1dB doesn't care about the types of harmonics created by the saturation, because as you mentioned, it's highly dependent on the types of circuits, but it's a measurement that could easily be used for all amplifiers because it defines the point of nonlinearity that can be used to determine the linear range for any amplifier. I think the problem is that most audio devices fudge their headroom numbers a little, and using compression point would probably dispel some of the marketing fluff they rely on. We call the point where we run out of voltage headroom "clipping". Strictly speaking it's not really the same thing as "saturation", although it's a related phenomenon. If one wishes to get nitpicky, saturation refers to the point in transformers/inductors where the core can no longer handle an increase in magnetic field; it becomes magnetically saturated, like a sponge that can no longer absorb additional water. It's not when the "water" supply runs out. In electronics, it's also called saturation. The transistor junctions are saturating with maximum amounts of current from lack of potential. AKA: it can't amplify any further because it's maxed out.
|
|
|
Post by mrholmes on Apr 3, 2020 10:49:38 GMT -6
Some noise can be nice.... With the years passing by....I have my problems with all those gain staging theories....
Pushing the input like crazy in context A can sound wonderful, abusing the same gear in context B sounds awful. A wise man on this MB told me "Use your ears" ( svart )
Note that pushing the input to your preamp/analog input chain does NOT mean that you need to push the input to the converter. There are these things called "output level controls" (aka "faders" in some contexts.) They're there for a reason.
My fault I should read original posts to an end... Remember a discussion with wiz (four years ago) since then I hit the input of my converters at low levels. IMO it sounds better. I just don't know why!!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 3, 2020 13:54:58 GMT -6
Intersample peaks are a very big deal John but no converter has a noisefloor of -144 dB. The analog parts of it don’t. The switchers in the chip don’t. The mics and mic pres don’t. You’re hitting resistor and switcher electrical noise. And all of that stack up of external and gear noise is just being converted and saved to the file in the computer. The converter outputs it as a 24-bit file but it itself has a lesser dynamic range. A cheap, crappy converter like a Focusrite Scarlet, Presonus, or Steinberg not the AXR4 thing, can’t even hit 16-bit THD+N. A workman like Babyface Pro will measure slightly cleaner than redbook standard. Now that’s just with a 1 kHz sine wave. When used with material that isn’t fixed sine waves, all bets are off. A Babyface Pro sure as hell isn’t that clean in the high end. The RME treble (not as bad as before but still bad) won’t allow that. Stack a bunch of tracks through it and the noise and distortion compound. Hence the constant search for cleaner and more accurate conversion or the acquiescence to conversion that distorts in a pleasing way like the Lavry Saturation and Burl transformers. We're talking anbout dBfs. That's a digital measurement; it has no bearing in the analog realm. A well designed converter should have an analog section that can deal with that. And what you furnish to the ME these days is a digital file; it shouldn't be affected by the analog output stage of your converter. The file that leaves the ME won't (or shouldn't) be at the same level as the input - part of mastering includes optimizing the output level for the delivery platform. I’m talking about input headroom. Worse converter makes a worse digital file with more noise despite both making 24 bit wav. Obviously in today’s world of 32 bit or better floating point daw, you can lower digital faders with hardly any detail loss but low levels going in mean less detail starting out and higher noise floors that come out when raising faders, compressing, or summing tracks. That wasn’t the case with 24-bit fixed point resampling like Yamaha O2R or Protools HD box. You still want as high levels as you can get without distorting or saturating the analog stages of the converter (unless it is one of the few that does this in a cool way) or intersample clipping. mastering engineers can jsut digitally lower faders of relatively uncompressed mixes if the peaks reach that high due to little to no bus compression. The problem is loudness war, limited mixes with heavy look ahead limiting of tracks and buses, guys intentionally clipping the ad to make stuff like snares stand out from limited drum or mix buses (I think this is retarded practice. Metallica seems to have done this with rhythm guitar on death magnetic cd vs guitar hero earlier mix but I have seen this mostly on modern “metal” snares and vocals), the bus compressor attacking too fast at too low threshold at too high ratios which can clamp down on all transients and acting almost as a soft limiter, and people sending buses and otherwise dynamic tracks through distorted tube gear (some of which is very well recommended, very expensive, but still utter shit for many genres imo) making the waveforms into a blob totally inappropriate for the genre and vibe of the recording. Mastering engineers want to commit all of these sins themselves and are mad at mixers for tying their hands, especially when they know the record will be compressed and limited to hell and want to at least their mix largely preserved.
|
|
|
Post by notneeson on Apr 3, 2020 14:44:48 GMT -6
Not my experience that you have to “track hot” to Pro Tools HD, by which I assume you mean the original 192s. Maybe you meant something else.
|
|
|
Post by johneppstein on Apr 3, 2020 16:22:40 GMT -6
We're talking anbout dBfs. That's a digital measurement; it has no bearing in the analog realm. A well designed converter should have an analog section that can deal with that. And what you furnish to the ME these days is a digital file; it shouldn't be affected by the analog output stage of your converter. The file that leaves the ME won't (or shouldn't) be at the same level as the input - part of mastering includes optimizing the output level for the delivery platform. I’m talking about input headroom. Worse converter makes a worse digital file with more noise despite both making 24 bit wav. Obviously in today’s world of 32 bit or better floating point daw, you can lower digital faders with hardly any detail loss but low levels going in mean less detail starting out and higher noise floors that come out when raising faders, compressing, or summing tracks. We're talking about digital levels, not input levels. And if your converter is such a piece of crap that it can't handle a +4dBu input without overdriving the digital stages... well, there's really no excuse.You need a converter that isn't junk. And all that's beside the point, anyway - turn down the output attenuator on your preamp.
You want to maintain good s/n in the analog stages, yeah. That does not mean that you should track so hot that you create intersample clipping in the mix. Which means tracking at a level of -15dBfs unless you only have very, very few channels in your mix* - which is something that you almost certainly can't be certain of during the initial tracking stages.
Once again - you gain ABSOLUTELY NOTHING by recording tracks at higher digital levels. Most converter drivers include a screen where you can adjust your digital level relative to the analog coming in. It's there for a reason. And that reason is NOT to crank everything up - it's there so you can achieve correct gain structure.
You want to overdrive something, overdrive your preamp. Or compressor, if you use one in front of your conversion. Then turn the output level down to a reasonable level. If you're using a comp that lacks an output trim, use it somewhere else. Or have a techie friend build you a simple passive attenuator box to tame it.
No, they can't "just digitally lower faders. The intersample clipping is created when you add channels in the mix stage. Once you do that it's done - no amount of anything you do afterwards (in the mastering stage) can undo it.
This is not a matter of tying their hands. It's a matter of delivering a damaged mix with flaws that cannot be compensated for - it has nothing to do with "tying their hands" in an artistic sense, which is something I tend to support, BTW.
I have exactly two words for such monkeys - "YOU'RE FIRED!"
If I deal with an ME I expect there to be a dialog about just what I'm after. And I expect to receive test recordings back for me to critique.
My mastering engineer would never dream of taking such liberties - and he's been in the game since the mid/late '60s and has mastered countless hits.
When I get an album back from mastering I expect it to follow my vision for the record EXACTLY. That does not mean that I don't want necessary adjustments made or have problems corrected that I couldn't figure out.
(Considerable digression about incompetent engineers in particular genres deleted)
* - like, maybe 5 or 6...
|
|
|
Post by Ward on Apr 3, 2020 16:27:05 GMT -6
What is the proper way to gain stage an LA-2A? I leave it in the chain with the reduction all the way down and the gain at 40 to start. Come off the pre, go directly into the console or converters. Average output should be at around -18dbFS. Which is 24 bit digital, unless they've changed it since 2019. This ensures you won't overload the next unit in your chain. Next, Break the chain and come out of the pre into the LA2a, hoping you're doing this with a good TT patchbay. At this point gain on 40 and peak reduction on 40 should give you very little compression and of course, you are in 'compress' mode, not 'limit' mode. Now as you increase gain reduction, compression, you'll also need to increase make-uo gain as well. Usually, if I'm averaging -18 into the LA2a, then 43 on gain and 45 on peak reduction gives me what I want.
|
|
|
Post by chessparov on Apr 3, 2020 16:29:53 GMT -6
So on low tracks and/or minimal Processing... -10 or -8 are always a good max level? Thanks, Chris
|
|
|
Post by tasteliketape on Apr 3, 2020 16:39:24 GMT -6
|
|
|
Post by popmann on Apr 3, 2020 18:07:47 GMT -6
I think Babel point here is using words like "level" as if the VU "level" is even related to the digital peak level.
There's no "right answer" for a digital peak level. There absolutely IS a correct answer for the VU level you should be aiming at...I looked at the digital stats because OP asked, but I set that level without any looking at a digital peak meter. So, it's not that I'm "aiming" for a digital peak of -14 on a vocal. In fact, later I futzed with the VCA a little, and the next take was -12.4dbfs peak....same VU reading, which is what I'm looking at, along with the gain reduction LEDs on the 661 to see when it's kicking in...point being, the digital peak isn't useful, except on really percussive close mic'd stuff where there's some chance of going over. If you tried to set a snare to healthy VU readings at -18 calibration, you would be over or you'd be walking it pretty close depending on whether you have an analog limiter patched in or not.
|
|
|
Post by Johnkenn on Apr 3, 2020 19:41:29 GMT -6
Switched to +24 on the Apollo. Gives me more play with the input gain of the pre...and popmann the VU on the insert was a killer idea. Now I don’t feel like I’ve got to do any translation.
|
|
|
Post by Ward on Apr 4, 2020 5:32:19 GMT -6
I weep for the razor industry.
|
|
|
Post by mrholmes on Apr 4, 2020 9:00:14 GMT -6
Just for clarity...I understand how input and output works on a mic pre (I’ve been doing this for a little while).. .maybe I should have asked how hot you’re going into the next stage...in order not to achieve distortion or clipping. I guess using my ears has served me well for several decades.
Do you remember the thread where some folks helped me to figure out that zero VU is the same in and out the box. With this I always have reference point how hard I hit the next unit. The advice to use ears is still valid some gear likes some more volts some lesser.... IMO
|
|
|
Post by chessparov on Apr 6, 2020 1:13:29 GMT -6
I weep for the razor industry. (Slap!) "Thanks, I needed that". Chris
|
|